A major challenge in human immunology is to develop tools for the unbiased analysis of immunity using low input samples from hundreds or thousands of patients. Current approaches have several limitations. First, tissue staining and flow cytometry are hard to scale up and only measure a small number of markers. Second, most genome/proteome-wide studies have focused on aggregated signals, such as mixtures of cells (PBMCs, whole tissues, tumors) that miss cell-type specific signals, or fluids, such as plasma, that combine secretions from billions of cells. Third, although unbiased studies of gene expression in purified immune cell types have shown success in discovery of disease predictors, it remains impractical to scale up such studies using conventional approaches for cell isolation such as flow cytometry or magnetic beads. What we urgently need are methods for unbiased profiling of purified cell types, which scale to large-scale human studies with low input samples. We propose to integrate two rapidly developing technologies -- microfluidics and sensitive RNA amplification -- into a compact device for highly multiplexed purification and RNA amplification of immune cell subsets. Such a major advance in unbiased measurements would advance our understanding of mechanisms and discovery of predictive markers for immune diseases. In preliminary data supporting this proposal, the Blainey lab has demonstrated a microfluidic device that can purify human immune cells using magnetic bead- based affinity purification, and separately, that can synthesize libraries for next generation sequencing (NGS). In parallel efforts, the Hacohen lab has developed low input RNA-seq protocols that work well on flow sorted immune cell types. These combined efforts support feasibility of a microfluidic system that couples cell purification with NGS library synthesis and enables sampling of all the major immune cell types from 1-5cc of blood. We propose to develop a microfluidics device to: (i) isolate immune cell subtypes directly from a mixture of cells; (ii) lyse purified cells; (iii) synthesize libraries for RNA-seq. We will demonstrate the utility of this technology in the analysis of primary human PBMCs from the blood of healthy subjects. Future applications to the clinic will focus, for example, on cross-sectional and longitudinal studies of autoimmune, infectious diseases, asthma, transplant and aging. Such studies are anticipated to lead to better understanding of immune responses and improvements in diagnosis, prognosis and therapeutics for human immune diseases. By creating the first scalable tools for unbiased analysis of cell states within the immune system, we hope to provide human immunologists the tools to study immune diseases at unprecedented depth.